CLI Commands

sse

Main processing command for extracting data from audio files.

Usage

sse <config_path> [options]

Arguments

config_path (required)

• Path to your .json configuration file

Options

-c, --cpu

• Disable GPU acceleration and force CPU-only processing
• Useful when GPU memory is insufficient

-m, --memory MEMORY

• Set maximum RAM limit in MB (Unix/macOS only)
• Does not limit GPU VRAM
• Example: -m 8192 limits to 8GB RAM

--validate

• Validate configuration file and exit
• Does not start processing or show interactive menu
• Displays configuration summary and validation status

--auto

• Automatically run all processing steps without interactive menu
• Useful for scripting and batch processing

-h, --help

• Show help message and exit

Examples

Interactive mode (default)

sse /path/to/config.json

Validate configuration only

sse /path/to/config.json --validate

Automatic processing without menu

sse /path/to/config.json --auto

Force CPU processing with memory limit

sse /path/to/config.json --cpu --memory 8192

---

sse_install_cuda

Download optional dependencies to use CUDA acceleration.

Usage

sse_install_cuda

warning

TensorFlow officially supports Linux and Windows WSL2 only.

Refer to official documentation for additional details.

---

sse_download

Download neural network models and resources for offline use.

Usage

sse_download

Downloads all required models:

• VGGish
• PERCH
• SurfPERCH
• YAMNet
• BirdNET
• Music Classifier

Output confirms when download is complete and system is ready for offline use.

tip

This command runs automatically when using our installation scripts.